1. Field
The invention relates to self-optimizing method and machine; and more particularly to multivariable, real-time self-optimizing method and machine, for operation in practical or nonideal conditions involving multiple errors due to defective knowledge bases, miscalibrated sensors, imperfect equipment, and statistical fluctuations.
2. Prior Art
Computer-controlled automation systems are widely used. All these systems are extremely powerful. With their uses, consistency is generally achieved together with the usually, but not necessarily always, associated improved profits, productivity, and product or service qualities (P.sup.3 Q).
Further, by transferring to machines human intelligence, rather than human skill, these systems have ushered us into this Second Industrial Revolution.
But human intelligence or knowledge bases only document averaged previous test results on old samples with equipment, materials, parts, procedures, or environment different from the present or future. Inherent are errors due to the various samplings, assumptions, or extrapolations, and repeated human interactions. These knowledge bases are often incomplete, inaccurate, biased, erroneous, out-of-date, and too generalized for uses in a particular automation task with a specific combination of equipment, procedures, materials, parts, and environment.
Detailed knowledge bases on modern technologies are particularly lacking. Even many old yet important technologies are not fully mastered or understood. Each of these technologies involves many processing steps often with unknown chemical, mechanical, electromagnetic, aerodynamic, fluidic, or other phenomena on the subatomic, atomic, microscopic, macroscopic, or other levels. Even the controlling variables in each phenomenon are often not completely known, and certainly have not been comprehensively studied. In fact, much R&D remains to be done in every field. Yet time is running very short in this highly competitive world. A new R&D method must, therefore, be developed.
For example, system dynamics of modern processes and equipment are generally ill-defined. Chemical and metallurgical reactions are often uncertain. The mechanism of catalysis is not completely known. Crystal growth is still an art. After millions of controlled welding experiments, the critical welding variables cannot yet be even identified among the very many possible. In the case of the new high temperature ceramic superconductors, the samples are still hard to make, shape, purify, reproduce, isolate, stabilize, confirm,, or even determine compositions.
Without reliable knowledge bases, the usual automation specialists would be at a loss in selecting the few among many manufacturing or servicing phenomena or variables to be controlled, in formulating the system dynamics models, in setting up the control equations, in determining the control constants, and in specifying setpoints for the control variables.
The fragile and unreliable knowledge bases often give only partial or invalid system dynamics models, oversimplified control equations, and inexact or misleading control constants and setpoints. In addition, all the setpoints are too arbitrary and round-numbered (e.g., 800.degree. C. and not 796.768.degree. C., 3 feet per minute, 20 gallons) to be possibly optimal statistically. The chance of these variables or setpoints being optimal at any time, not to say instantaneously or continuously, is nearly zero. Further, the optimal setpoints cannot be constant, as are assumed in present automation systems, but change with variations in time, equipment, procedures, materials, parts, and environment.
These conventional automation systems are also not smart and must be spoon-fed at every step via computer programs or master-slave instructions. They are not totally integrated or automated, and require constant human guidance, reviews, analyses, interactions, and supervision.
Due to this repeated human involvement, the conventional automation systems not only are greatly slowed down, but inevitably inherit the many defects of the inconsistent and imperfect human test planners, samplers, testers, data collectors and analyzers, communicators, and technicians. Humans are million times slower and less reliable than microprocessors in, e.g., memory recalling or storing, information inputting or outputting, data analyzing, communicating, and commanding or actuating.
In addition, usually these present systems merely passively adapt, adjust, correct, control, or regulate, in response to variations in the environment or a few control variables. Dealing with more than several interacting variables results in extremely large number of tests to be made; and in massive amount of data to be collected, conditioned, stored, and quickly or instantly analyzed. This is often impractical or impossible, because of the well-known problems of "combinatorial explosion" and "computer intractability," as has been described in the '864 patent.
Yet, modern technologies invariably involve many unpredictable, interacting, and rapidly changing control variables in such categories as: raw materials, vendors, batches, lots, and conditions; compositioning; processing equipment; procedures in each of the many steps in every process; and environment. Many phenomena are transient but highly nonreproducible yet unknown and critical.
Artificial intelligence (AI) technologies, particularly the expert systems, have been developed and increasingly used in various fields. But again the knowledge bases are often inadequate or deficient, particularly on developing technologies. The present expert systems are also costly, inflexible, qualitative, and often inaccurate and out-of-date particularly for complicated yet rapidly improving modern technologies. In addition, they too cannot handle the inherent large number of interacting variables.
Reliable and relevant knowledge is scarce and very costly. Up to now, the main bottleneck in the development of expert systems has been the acquiring of the knowledge in computer-usable form. Human knowledge often not only is fragile, costly, unreliable, but difficult to be translated for uses by machines. Codifying an expert's skill has always been a long and labor-intensive process.
Hence, experts conclude that machine learning is the key to the future of automation in general and expert systems in particular. The valuable knowledge must be manufactured, in bulk and at low cost. So far, however, no such machines exist.
Conventional AI development environments have difficulties in producing efficient real-time systems. This is due to the fact that the same code necessary to enhance the development environment tends to slow down the system during run-time. To overcome these limitations, AI system designers must embed the knowledge base (KB) into their own custom run-time AI shells to achieve real-time performance. Unfortunately, the deeper the KB is embedded into the actual code, the harder it is to change the KB when maintenance is necessary. Therefore, the AI system designer must constantly balance system performance versus ease of maintaining and manipulating the KB. An automation system with real-time KB generating capacity would thus be highly desirable.
Present automation systems also invariably contain various hidden errors of samplings, assumptions, extrapolations, scaling-ups, and statistical fluctuations of uncertain magnitudes. These systems are also at the mercy of other errors due to, e.g., miscalibrated sensors, imperfect actuators, drifting equipment, and partially damaged components. Any one of these errors can easily lead to unexpected inconsistencies and nonuniformities in, e.g., manufacturing or servicing results.
Accordingly, an object of the present invention is to provide improved self-optimizing machine and method;
A further object of the invention is to provide real-time self-optimizing machine and method capable of handling tens, hundreds, thousands, or more variables with little or minimum human guidance;
Another object of this invention is to provide close-loop, self-optimizing machine or method which can optimize practically continuously and instantly;
A broad object of the invention is to provide self-optimizing machine or method totally self-planned and controlled tests, performed on the very particular machine or method itself without relying on many assumptions, invalid scaling-up laws, and extrapolation from sampled test results obtained on other similar machines or methods; with the test data instantly analyzed for timely optimization results;
Another object of the invention is to self-optimize machine or method in practical or nonideal conditions by tolerating, neutralizing, or suppressing the effect of errors due to defective knowledge bases, miscalibrated sensors, imperfect actuators, drifting equipment, damaged components, and statistical fluctuations;
A further object of the invention is to provide self-optimizing machine or method which operates with deficient and minimal or practically zero knowledge bases, rapidly generates its own new knowledge bases through automatic R&D, and immediately and continuously replaces these new knowledge bases with still newer and more accurate or refined knowledge bases for continuously optimal results;
An additional object of the invention is to provide self-optimizing machine and method which actively computes, and automatically sets at, the instantaneous optimal combinations of the many relevant variables in various categories, with instant feed-back to supply data for immediate replanning, retesting, and reoptimiizng, all without human intervention;
Another object of the invention is to manufacture, in bulk and at low cost, reliable knowledge bases for the instant development of relevant expert systems;
Yet another object of the invention is to provide self-optimized computer system software and service for businesses, offices, education, training, engineering, reverse-engineering, designing, processing, manufacturing, distribution, R&D, communication, reconnaissance, surveillance, project management, data analyses, decision making, running supercomputers or parallel computers, multimedia computing, electronic data interchanging, computer operation and application, and other applications;
A further object of the invention is to provide computer-generated software or creatures which, all by themselves and with blinding speed, reproduce or multiply, mutate or undergo spontaneous genetic changes in the best possible way under a given environment.
Further objects and advantages of my invention will appear as the specification proceeds.